CopperPilot, a startup building an AI design agent for hardware engineers, has raised $390,000 from Neon Fund, alongside investments from US-based angel investors.
Founded in October 2025 by Aditya NG, Mohit Bagur, and Anish Harsha, the startup is building an AI agent that converts plain-English descriptions into manufacturing-ready circuit board designs.
The company is also developing software that integrates AI with existing hardware design tools to automate tasks such as schematic capture, PCB layout, routing, simulation and firmware development.
The funds will fund product research and engineering, improve the agent’s performance and reliability, and expand the engineering team.
The company’s active users have grown 25-fold since August, with more than 6,000 design artefacts run on the platform to date, says Aditya, Co-founder and CEO of CopperPilot, in an exclusive interview with AIM.
CopperPilot has also opened the first version of its design agent, codenamed Epsilon, to hardware engineers. “We are hardware engineers ourselves. This is a tool that we deeply wanted,” Aditya says.
His own experience in autonomous vehicles helped shape CopperPilot’s founding thesis. Aditya has worked on autonomous vehicle projects at Purdue University and Minus Zero.
He says hardware development had prevented teams from pursuing projects that were technically possible but too difficult to execute with limited resources. CopperPilot’s long-term goal is to make such projects more accessible to smaller teams by reducing the engineering effort required to build and validate hardware.
“You should never be bottlenecked because it was too hard to design something,” Aditya says. “Much like today, you’re not blocked if you want to design a piece of software. You should not be blocked when you want to design a piece of hardware.”
Bringing AI into Existing Hardware Design Tools
Hardware engineers typically work across multiple tools and documents to develop a product. They must read component datasheets, understand manufacturer specifications, create circuit schematics, lay out printed circuit boards (PCBs), check design rules and validate the design before manufacturing.
CopperPilot seeks to bring these activities into a single workflow by integrating with existing design software and firmware frameworks rather than asking engineers to migrate to a new platform.
“Instead of asking them to migrate away from that design tool, because that’s a lot of legacy software and legacy designs, we bring the AI to them,” Aditya says.
The company supports tools including Altium, KiCad, OrCAD, and EasyEDA, along with simulation software such as LTspice and firmware development environments. It also integrates with AI coding tools such as Claude Code, Cursor, and Codex, letting developers start hardware design tasks within these applications.
CopperPilot supports the entire engineering process, from ideation and schematic capture to PCB layout, routing, design-for-manufacturing checks, and, in some cases, firmware development.
The goal is to reduce the manual work of implementing and checking designs, letting engineers focus on product specifications and system architecture.
Using Datasheets and Simulations to Validate Designs
One challenge in applying AI to hardware engineering is ensuring generated circuits are compatible with real components and physical constraints. A design must account for component specifications, electrical requirements, and manufacturing limitations, not just produce a plausible circuit.
CopperPilot addresses this by extracting information from component datasheets and manufacturer documentation and incorporating those constraints into the design process. The agent can also interact with simulation tools to test parts of a circuit before presenting its findings to the engineer.
“We’re trying to mimic how a hardware engineer would actually go about designing,” Aditya says, describing how the agent reads datasheets, consults manufacturer information, builds a design and simulates it where applicable.
The agent can also identify infeasible user specifications and seek clarification before proceeding. During validation, it can report which test cases it has covered and how confident it is in the design.
CopperPilot uses third-party AI models, selecting what it considers the best available option for its requirements. Aditya says the company could fine-tune its own model, adopt an open-source model or use a model from a major AI provider as its needs evolve.
Engineers Retain Control Over the Design
CopperPilot can automate several stages of hardware development, but it does not seek to remove engineers from the process. The company keeps users involved in decisions about specifications, architecture and the intended outcome.
For a straightforward task, such as replacing a full-sized HDMI port with a micro-HDMI port, an engineer could delegate the change and review the completed design. A more complex task, such as developing a motherboard, would require more interaction to clarify requirements.
Engineers can monitor the agent in their existing design software, interrupt it if needed, make changes themselves, and ask it to continue with those corrections.
“The important thing we still want the user on is the taste, the design specification, and the architecture,” Aditya says.
This approach also extends to debugging. For example, if a device shuts down unexpectedly, CopperPilot can examine its PCB design, firmware, logs and other available inputs to investigate the cause.
The agent could identify a mismatch between a power supply’s current rating and the power drawn when a machine learning model activates a processor or GPU. It could then suggest changes, such as using a different power supply or adding capacitors to handle current spikes.
Aditya says the ability to connect information across design tools helps the agent identify problems that an experienced engineer might otherwise have to trace manually.
The startup said designing electronics, from creating circuit diagrams to ordering the first prototype boards, can take six to 15 weeks and cost between $7,500 and $30,000. CopperPilot says its platform can reduce this process to a few hours.
Targeting Hardware Companies
CopperPilot is targeting industries including consumer electronics, renewable energy, energy storage, robotics, the Internet of Things (IoT), automotive and electric vehicles.
The founders spent about a year developing and benchmarking the design agent, drawing on their experience in autonomous vehicles, quantum photonics and robotics.
The company has customers in the US, UK, and Europe, as well as users in India and other markets.
Aditya says the company is seeking to build adoption among individual engineers who can advocate for the tool within their organisations, as well as founders and managers who can approve wider deployment.
